Salivette, a relevant saliva sampling device for SARS-CoV-2 detection

Figures & data

Table 1. Characteristics of participants under investigation for COVID-19 diagnosis by paired NPSs and saliva samples

	Overall	Follow-up*	Symptomatic	Asymptomatic
Participants, n (%)	303	30 (9.9%)	90 (29.7%)	183 (60.4%)
Age (years), median (IQR)	40 (29–50)	45 (34.5–58.8)	41 (28.3–50.0)	39 (28.5–49.5)
Male, n (%)	144 (47.5%)	16 (53.3%)	35 (38.9%)	93 (50.8%)
Onset of symptoms before the test (days), median (IQR)	0 (0–2.5)	8 (6–11)	3 (1–5)	/
Symptoms at presentation, n (%)	104 (34.3%)	14 (46.7%)	90 (100%)	0 (0.0%)
Cough, n (%)	21 (6.9%)	3 (10.0%)	18 (20.0%)	/
Sore throat, n (%)	16 (5.3%)	1 (3.3%)	15 (16.7%)	/
Runny nose, n (%)	18 (5.9%)	0 (0.0%)	18 (20.0%)	/
Anosmia/Ageusia, n (%)	20 (6.6%)	6 (20.0%)	14 (15.6%)	/
Diarrhea, n (%)	6 (2.0%)	0 (0.0%)	6 (6.7%)	/
Myalgia, n (%)	31 (10.3%)	3 (10.0%)	28 (31.1%)	/
Fever, n (%)	39 (12.9%)	3 (10.0%)	36 (40.0%)	/
Tiredness, n (%)	12 (4.0%)	3 (10.0%)	9 (10.0%)	/
Headache, n (%)	24 (7.9%)	0 (0.0%)	24 (26.7%)	/
Others, n (%)	3 (1.0%)	1 (3.3%)	2 (2.2%)	/

* Previously tested positively for SARS-CoV-2 by RT-qPCR on NPSs. Abbreviations: IQR, interquartile range; NPS, nasopharyngeal swab; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; SD, standard deviation.

Table 2. Comparison of the RT-qPCR detection of SARS-CoV-2 between NPSs and saliva samples

Saliva samples	NPSs		Total
	Positive	Negative
Positive	29	23	52
Negative	6	245	251
Total	35	268	303
Agreement (%)	90.4%
Cohen’s κ ^#	0.652 (Substantial)
Sensitivity (%)	82.9%
Specificity (%)	91.4%

^#Coefficient of agreement, the agreement level is indicated into brackets, as previously defined [35]. NPS, nasopharyngeal swab.

Figure 1. Comparison of Ct values from NPSs and saliva samples. (A) All positive NPSs (n = 35) and saliva samples (n = 52) were compared using a Mann–Whitney test (p = 0.097). Bars represent the median and 95% CI. (B) Paired positive samples (n = 29), represented by the connecting lines, were compared by a Wilcoxon test (p = 0.761)

Table 3. Comparison of the RT-qPCR detection of SARS-CoV-2 between NPSs and saliva samples according to participant status

Saliva samples	NPSs						Total
	Follow-up*		Symptomatic		Asymptomatic
	Positive	Negative	Positive	Negative	Positive	Negative
Positive	2	6	20	7	7	10	52
Negative	4	18	2	61	0	166	251
Total	6	24	22	68	7	176	303
Agreement (%)	66.7%		90.0%		94.5%
Cohen’s κ ^#	0.074 (Slight)		0.749 (Substantial)		0.559 (Moderate)
Sensitivity (%)	33.3%		90.9%		100%
Specificity (%)	75.0%		89.7%		94.3%

*Previously tested positively for SARS-CoV-2 by RT-qPCR on NPSs. ^#Coefficient of agreement, the agreement level is indicated into brackets, as previously defined [35]. NPS, nasopharyngeal swab.

Table 4. Comparison of the RT-qPCR detection of SARS-CoV-2 between NPSs and saliva samples to a reference that considers a person to be positive if one of his or her samples is positive

	NPSs		Saliva		Total	Chi-square test	95% CI
	Positive	Negative	Positive	Negative
Positive	35	0	52	0
Negative	23	245	6	245
Total	58	245	58	268	303
Agreement (%)	92.4%		98.0%			p < 0.003	[−0.093; −0.019]
Cohen’s κ ^#	0.711 (Substantial)		0.933 (Almost perfect)
PPA (%)	60.3%		89.7%			p < 0.001	[−0.458; −0.128]

^#Coefficient of agreement, the agreement level is indicated into brackets, as previously defined [35]. CI, confidence interval; NPS, nasopharyngeal swab; PPA, positive percent agreement.

Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–174.

 PubMed Web of Science ®Google Scholar